There seems to be considerable interest in which bridge option will be chosen. The three initial options were an arch bridge, a cable-stayed bridge (shown top), and a structure described as a "wave frame" (shown bottom) but which is essentially a somewhat dubious cross between a Vierendeel truss and a suspension bridge. Of those, they seem to have whittled the choice down to several variations on the cable-stayed and "wave frame" options.
The somewhat unconventional "wave" option seems popular with the natives, if a poll at BlueOregon is to be believed. The Portland Spaces blog is sufficiently keen that they interview the designer, architect Miguel Rosales.
You might think that Rosales would know his stuff: he worked on the US$115m cable-stayed Leonard P. Zakim Bunker Hill Bridge in Boston, as well as the Liberty Bridge in Greenville, South Carolina, a curved suspension bridge. It may seem somewhat odd, however, that his website doesn't so much as mention the engineers who played key roles in these structures (Christian Menn in Boston - who returns the favour by making no mention of Rosales; and Schlaich Bergermann in Greenville).
Reading the interview with Rosales at Portland Spaces, there's still no mention of an engineer, and I came away with the distinct impression that this is a bridge promoted by an architect who doesn't really understand bridge design. In the comments at the foot of the interview, Rosales says:
"The flowing top chord follows the flow of forces and moment diagram resulting in a slender and elongated structure. The clean design avoids the visual confusion often found in steel trusses with their multiple layers of diagonal and vertical members."But of course, it doesn't follow the moment diagram at all, which would be cusped above the piers (i.e. "pointy"), in the same manner as a conventional suspension bridge. And it avoids the truss diagonals only at the expense of requiring massive amounts of additional steelwork to provide equivalent stiffness (and it's stiffness that's normally critical on a light-rail bridge).
It's far from clear how the bridge (which spans about 200m) actually works - the very slender depth at midspan provides insufficient moment continuity for it to work as an efficient continous truss, while the weedy vertical members lack the stiffness required for suspension bridge towers. If, as it appears, it's a bizarre hybrid self-anchored suspension bridge relying on Vierendeel action, throughout, it's no surprise that TriMet's cost estimates show it as significantly more expensive than the cable-stayed alternative.
Rosales is keen to promote his design, and suggests that TriMet's engineers have been unreasonable with their cost estimates. From the TriMet website [PDF], it's apparent they have put considerable effort into trying to make the Rosales design work, but still estimate its cost at US$119m, against US$93m for the cable-stay option. That works out at about US$10.5k per square metre of deck for the "wave frame", or about £7,000. That seems about right for an unconventional design like this, but it's easy to suggest it could be more.
Quick! Is there an engineer in the house?